Botulism is caused by Clostridium botulinum neurotoxin (BoNT), a CDC Category A biodefense threat agent for which no antidote exists. BoNT is the most potent toxin known and can be produced with relative ease. If large numbers of people were exposed to even a small dose of this toxin, they would become paralyzed and require assisted breathing which could easily overwhelm the limited supplies of respirators. We have developed small protein agents that, when expressed within intoxicated neurons, promote rapid degradation of BoNT proteases and improve rates of recovery from intoxication. We call these biomolecule agents targeted F-boxes (TFBs) because they consist of a 15 kDa F-box domain fused to a 14 kDa camelid VHH domain with binding specificity for a BoNT protease. Here we propose to develop adenovirus-based botulism antidotes that will specifically target motor neurons and lead to cytosolic expression of a TFB agent within intoxicated neurons. Adenoviruses (Ads) are ideal vehicles for delivery of TFBs to motor neurons for several reasons. First, Ad does not integrate their genome into infected cells and Ad vectors have been produced that are non-replicating and safe for therapeutic use. Second, Ad can be modified to dramatically reduce the normal tropism (liver) while increasing infection of a selected cell population. Thirdly, Ad vectors can be modified to produce a transient burst of transgene expression lasting only a few weeks after which the infected cells revert to normal. Finally, Ads have been found to infect motor neurons leading to transgene expression and, surprisingly, BoNT intoxicated motor neurons are much more efficiently infected than normal neurons. In the R21 phase of this proposed project, we will develop an adapter protein that will block normal Ad tropism while imparting tropism for motor neurons by employing a neuron-specific receptor binding domain (RBD) that derives from BoNT. Recombinant Ad will be engineered to express a TFB agent that targets the protease from BoNT serotype A (BoNT/A) for degradation and will be pre-treated with the neuron-targeting adapter. This modified Ad will then be tested as a botulism antidote by local injection into BoNT/A intoxicated muscles in mice. If successful, in the R33 phase we will make additional modifications to the Ad vector to further improve specificity for motor neurons and obviate the need for an adapter protein. We will also engineer an Ad antidote for another BoNT serotype, BoNT/B. These Ad vectors will be tested for the ability to accelerate recovery from botulism paralysis in mice following a systemic administration of the therapeutic agent. If successful, similar agents can be readily developed for all seven known BoNT serotypes by simply replacing the VHH domain of the TFBs with others having specificity for different BoNT proteases. In addition, the neuron-targeted Ad vehicles we develop may have applications in other important neuronal pathologies.